BEGIN:VCALENDAR
VERSION:2.0
PRODID:IEEE vTools.Events//EN
CALSCALE:GREGORIAN
BEGIN:VTIMEZONE
TZID:America/Los_Angeles
BEGIN:DAYLIGHT
DTSTART:20250309T030000
TZOFFSETFROM:-0800
TZOFFSETTO:-0700
RRULE:FREQ=YEARLY;BYDAY=2SU;BYMONTH=3
TZNAME:PDT
END:DAYLIGHT
BEGIN:STANDARD
DTSTART:20251102T010000
TZOFFSETFROM:-0700
TZOFFSETTO:-0800
RRULE:FREQ=YEARLY;BYDAY=1SU;BYMONTH=11
TZNAME:PST
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20250827T013856Z
UID:DC279D30-3882-45F9-898F-A1216D82BE1C
DTSTART;TZID=America/Los_Angeles:20250826T173000
DTEND;TZID=America/Los_Angeles:20250826T183000
DESCRIPTION:Recent advancements in machine learning\, while powerful\, are 
 often burdened by significant computational and memory requirements\, limi
 ting their deployment in resource-constrained settings. Hyper dimensional 
 Computing (HDC) emerges as an alternative with its simplicity\, lightweigh
 t operations\, and robustness to errors. By encoding data into high-dimens
 ional vectors and performing efficient algebraic computations\, HDC opens 
 a new avenue as an efficient learning paradigm. In this talk\, Dr. Fatemeh
  Asgarinejad will introduce the fundamentals of HDC and briefly discusses 
 existing research that has extensively explored various stages of HDC algo
 rithm. Then\, she will present three key domains of her research: First\, 
 she will discuss PIONEER\, a novel approach that employs learned projectio
 n vectors to optimize the encoding process. By leveraging neural networks 
 to learn these vectors\, PIONEER enables HDC to achieve high accuracy with
  significant computational efficiency. Second\, she will present HDXpose\,
  an adversarial attack framework that exploits an advantage of HDC: “exp
 lainability”. By strategically analyzing and perturbing influential inpu
 t points\, HDXpose effectively unveils vulnerabilities within HDC models\,
  underscoring the need for robust security measures in HDC system design. 
 Lastly\, Dr. Asgarinejad will show an application of HDC in developing a c
 ost-effective and noise-resilient pressure mat system for human activity r
 ecognition. The HDC-based system surpasses CNNs in accuracy and efficiency
 .\n\nCo-sponsored by: Media Partner: Open Research Institute (ORI)\n\nSpea
 ker(s): Fatemeh Asgarinejad\n\nAgenda: \n- Invited talk from Dr. Fatemeh A
 sgarinejad\, an incoming Assistant Professor of Teaching in the Electrical
  and Computer Engineering Department at the University of California\, Riv
 erside.\n- Q/A Session\n\nVirtual: https://events.vtools.ieee.org/m/496707
LOCATION:Virtual: https://events.vtools.ieee.org/m/496707
ORGANIZER:upalmahbub@yahoo.com
SEQUENCE:46
SUMMARY:Towards Efficient Learning on Edge by Hyperdimensional Computing
URL;VALUE=URI:https://events.vtools.ieee.org/m/496707
X-ALT-DESC:Description: <br /><p>Recent advancements in machine learning\, 
 while powerful\, are often burdened by significant computational and memor
 y requirements\, limiting their deployment in resource-constrained setting
 s. Hyper dimensional Computing (HDC) emerges as an alternative with its si
 mplicity\, lightweight operations\, and robustness to errors. By encoding 
 data into high-dimensional vectors and performing efficient algebraic comp
 utations\, HDC opens a new avenue as an efficient learning paradigm. In th
 is talk\, Dr. Fatemeh Asgarinejad will introduce the fundamentals of HDC a
 nd briefly discusses existing research that has extensively explored vario
 us stages of HDC algorithm. Then\, she will present three key domains of h
 er research: First\, she will discuss PIONEER\, a novel approach that empl
 oys learned projection vectors to optimize the encoding process. By levera
 ging neural networks to learn these vectors\, PIONEER enables HDC to achie
 ve high accuracy with significant computational efficiency. Second\, she w
 ill present HDXpose\, an adversarial attack framework that exploits an adv
 antage of HDC: &ldquo\;explainability&rdquo\;. By strategically analyzing 
 and perturbing influential input points\, HDXpose effectively unveils vuln
 erabilities within HDC models\, underscoring the need for robust security 
 measures in HDC system design. Lastly\, Dr. Asgarinejad will show an appli
 cation of HDC in developing a cost-effective and noise-resilient pressure 
 mat system for human activity recognition. The HDC-based system surpasses 
 CNNs in accuracy and efficiency.</p><br /><br />Agenda: <br /><ul>\n<li>In
 vited talk from Dr. Fatemeh Asgarinejad\, an incoming Assistant Professor 
 of Teaching in the Electrical and&nbsp\;Computer Engineering Department at
  the University of California\, Riverside.</li>\n<li>Q/A Session</li>\n</u
 l>
END:VEVENT
END:VCALENDAR